Anxiety disorders are a risk factor for alcohol use disorders (AUDs), but mechanisms of risk are not well understood. One possibility is that alcohol is more negatively reinforcing for anxious individuals. Supporting this, studies show that anxious individuals report greater tension-reduction expectancies for alcohol and higher coping motives for drinking. Studies also suggest that consuming alcohol prior to a stressor may provide greater reduction of stress response for anxious individuals, but these results do not necessarily generalize to drinking in the absence of an imminent stressor (which likely characterizes most real-world drinking). The few studies conducted in the absence of an imminent stressor have failed to utilize adequate placebo controls or have not used validated measures of alcohol response. Studies meeting these criteria are needed to establish whether anxious individuals receive greater negative reinforcement from alcohol above and beyond expectancies. Existing studies also fail to address effects of environmental context, which has been shown to affect subjective response to alcohol (e.g., greater feelings of relaxation in less stimulating environments, stronger positive subjective effects in groups vs. solitary settings). If low-stimulation contexts facilitate greater negative reinforcement for anxious individuals, this would aid in identifying those at greatest risk for AUDs (e.g., anxious solitary drinkers). Finally understanding mechanisms underlying negatively reinforcing effects of alcohol in anxious individuals may lead to the identification of key targets for intervention and prevention of alcoho use disorders in this population. One possible mechanism is cortisol response to alcohol, which is blunted in heavy drinkers compared to light drinkers on the descending limb of the blood alcohol curve. Heavy drinkers also experience stronger positive and weaker negative subjective effects from alcohol compared to light drinkers (which is predictive of future alcohol use and problems). This suggests that cortisol response may influence subjective response to alcohol, at least on the descending limb, where these effects predominate. So, perhaps cortisol response is a marker of the valence of such low-arousal subjective effects (e.g., elevated cortisol = low-arousal effects experienced as negative, e.g., wooziness; blunted cortisol = low-arousal effects experienced as positive, e.g., relaxation). If cortisol response does influence subjective effects, anxious individuals may be at risk for developing alcohol problems via this mechanism because they, similar to heavy drinkers, show a dysregulated HPA axis due to chronic activation. The proposed research training plan will use a placebo-controlled alcohol administration design to investigate the following aims: 1) Determine whether anxiety symptoms are associated with a negatively reinforcing profile of subjective response relative to placebo. 2) Determine whether drinking context (physical and social) moderates the relationship between anxiety symptoms and alcohol response. 3) Investigate cortisol response to alcohol as a potential mediator of the relationship between anxiety symptoms and subjective response to alcohol.